Pneumococcal Virulence
Dissertation
zur Erlangung des akademisches Grades doctor rerum naturalium
(Dr. rer. nat.) im Fach Biologie eingereicht an der
Mathematisch-Naturwissenschaftliche Fakultät I der Humboldt-Universität zu Berlin
von
Diplom Biologe Florian Gehre geboren am 09.Juli 1978 in Erlangen
Präsident der Humboldt-Universität zu Berlin Prof. Dr. Dr. h.c. Christoph Markschies
Dekan der Mathematisch-Naturwissenschaftliche Fakultät I Prof. Dr. Lutz-Helmut Schön
Gutachter: 1. Prof. Dr Rainer Borriss, Berlin 2. Prof. Dr Alf Hamann, Berlin
3. Prof. Dr Alexander Tomasz, New York eingereicht: 5. Juni 2009
Datum der Promotion: 22. Oktober 2009
Streptococcus pneumoniae (the “pneumococcus”) is a gram-positive bacterium. It is a human pathogen that leads to severe diseases such as pneumonia, sepsis, otitis media and meningitis. Being a gram-positive bacterium, it is surrounded by a thick cell wall layer. This cell wall represents the outer surface of the pathogen and there- fore is the major target for the immune system during an infection. The aim of my thesis was to understand the impact of the cell wall structure on pneumococcal viru- lence. Particularly I was interested to test the role of acetyl-groups and choline resi- dues of the cell wall in the pathology of the disease. Therefore I studied deacetylated and choline-free pneumococcal mutants not only biochemically in vitro but also in respect to their virulence in various animal models of pneumococcal disease.
In the first part of my studies I analyzed a recently discovered mutant bacterium, de- ficient in gene adr, which was described to have decreased resistance to lysozyme in vitro. By DNA sequence comparison and chemical analysis of highly purified cell wall I was able to identify adr as the structural gene of the pneumococcal peptidoglycan O-acetyl-transferase. Since Adr is responsible to attach O-linked acetyl groups to the N-acetyl muramic acid residues of the cell wall, adr mutant bacteria lack this cell wall modification. I further demonstrated that Adr does not have any impact on pneumo- coccal attachment to human pharyngeal cells in vitro. However, adr mutant bacteria showed a dramatic decrease in their capacity to colonize the murine nasopharynx in vivo. This impairment is most likely due to their enhanced sensitivity to lysozyme.
In the second part of my thesis I worked on choline residues, a structural component of the (lipo)teichoic acids of the cell wall. This part of the thesis was especially intrigu- ing since S. pneumoniae has an auxotrophic requirement for this nutrient. Therefore, the recent construction of a choline-independent strain Cho- allowed me to investi- gate in detail the role that this aminoalcohol plays in the virulence of the pneumococ- cus during meningitis and sepsis.
For the meningitis model, the choline containing strain D39Cho- and its isogenic cho- line-free derivative D39Cho-licA64 (each expressing the capsule polysaccharide 2) were introduced intracisternally into 11 days old Wistar rats. During the first 8 h post infection both strains multiplied and stimulated a similar immune response that in- volved expression of high levels of proinflammatory cytokines, the matrix metallopro-
cerebro-spinal fluid (CSF). Virtually identical immune response was also elicited by intracisternal inoculation of either choline-containing or choline-free cell walls. At sampling times past 8 h strain D39Cho- continued to replicate accompanied by an intense inflammatory response and strong granulocytic pleiocytosis. Animals infected with D39Cho- died within 20 h and histopathology revealed brain damage in the cerebral cortex and hippocampus. In contrast, the initial immune response generated by the choline-free strain D39Cho-licA64 began to decline after the first 8 h accom- panied by elimination of the bacteria from the CSF in parallel with a strong WBC re- sponse peaking at 8 h after infection. All animals survived and there was no evidence for brain damage.
Using the same pair of strains in the murine sepsis model I demonstrated that this choline-associated virulence is independent of Toll-like receptor TLR-2 recognition.
Also, despite the lack of virulence, choline-free strains of S. pneumoniae were able to activate splenic dendritic cells, induce production of proinflammatory cytokines as well as capsule-specific serum antibodies and develop protective immunity against subsequent challenge with the virulent strain. However, due to this transient en- gagement of the immune system the choline-free bacteria were rapidly cleared from the blood while the isogenic virulent strain D39Cho- continued to grow accompanied by prolonged expression of cytokines eventually killing the experimental animals.
From the meningitis and sepsis model I was able to conclude that choline allows bac- teria to steadily grow within the host even in spite of an ongoing immune response.
Therefore attachment of choline represents a mechanism to evade host clearance. In the last part of my thesis I was able to show that surface bound choline residues can confer resistance against the bactericidal activities of complement-free murine serum ex vivo and the cationic antimicrobial peptide Nisin in vitro. Removing or blocking the choline-residues with either choline-specific IgA TEPC-15 antibodies or human C- reactive protein abolishes the observed resistance. Passive application of these two immune molecules protects mice from pneumococcal colonization and sepsis, sug- gesting a possible mode of interaction between IgA antibodies / CRP and cationic antimicrobial peptides in their fight against bacteria, especially S. pneumoniae.
Streptococcus pneumoniae ist ein gram-positives Bakterium und der Erreger von Lungen- sowie Mittelohrentzündung, Sepsis und Meningitis. S. pneumoniae ist um- hüllt von einer dicken Zellwand, welche ein Charakteristikum von gram-positiven Bakterien darstellt. Diese Zellwand bildet zugleich die physische Oberfläche der Zelle und ist deshalb eines der Hauptziele des Wirtsimmunsystems. Das Ziel meiner Dok- torarbeit war es zu verstehen inwiefern die Zellwandstruktur Auswirkungen auf die Virulenz des Erregers hat. Ich war vor allem an O-Acetyl-Seitengruppen und Cholin- resten der bakteriellen Zellwand interessiert. Deshalb analysierte ich zwei Bakterien- stämme, welche entweder vollkommen deacetyliert oder frei von Cholin waren.
Der erste Teil meiner Doktorarbeit beschäftigte sich mit einem vor Kurzem isolierten Bakterium, in welchem das Gen adr inaktiviert war. Es war bekannt, dass dieses Bakterium höchst Lysozym-sensitiv war. Von der DNA Sequenz des Gens und der biochemischen Analyse der Zellwand konnte ich darauf schliessen, dass jenes Gen adr für eine Peptidoglykan O-Acetyl-Transferase kodiert. Da dieses Enzym die N- Acetyl-Muraminsäure-Reste der Zellwand acetyliert, war diese Zellwandmodifikation in der entsprechenden adr Mutante nicht zu finden. Obwohl Acetylgruppen keinerlei Einfluss auf das Binden an humane Pharynxzellen in vitro hatten, zeigten adr Mutan- ten eine verminderte Fähigkeit den Nasopharynx von Mäusen zu besiedeln. Dies war auf die erhöhte Lysozym-Sensitivität zurückzuführen.
Im zweiten Teil meiner Doktorarbeit konzentrierte ich mich auf die Cholinreste der Zellwand. Diese Cholinreste sind Bestandteil der (Lipo)teichonsäuren der Zellwand.
Cholin stellt eine Besonderheit in der Physiologie von S. pneumoniae dar, da das Bakterium auxotroph für diesen Aminoalkohol ist. Cholin muss also ein obligater und essentieller Bestandteil des entsprechenden Wachtumsmediums sein. Unser Labor war jedoch in der Lage eine S. pneumoniae Mutante (Cho-) zu erstellen, welche fähig ist in cholinfreier Umgebung zu wachsen. Dank des Stammes Cho- konnte ich die Auswirkungen von Cholin auf die Virulenz des Bakteriums während experimenteller Pneumokokken-Meningitis und Sepsis testen.
Für das Meningitis-Tiermodell wurden 11 Tage alten Wistar Ratten entweder der cholinhaltige Stamm D39Cho- oder die cholinfreien D39Cho-licA64 Bakterien in die Cisterna magna injiziert. Während den folgenden 8 h waren beide Bakterien in der
Matrixmetalloproteinase 9 (MMP-9), Interleukin-10 sowie das Einströmen von weis- sen Blutkörperchen zu stimulieren. Interessanterweise konnte eine vergleichbare Immunantwort auch durch die Applikation von aufgereinigten cholinhaltigen und cho- linfreien Zellwänden ausgelöst werden. Nach der anfänglichen, achtstündigen Wachstumsphase waren nur die cholinhaltigen Bakterien in der Lage weiter zu wachsen. Dies führte zu einer stetigen Verstärkung der Immunantwort und dem Tod der Versuchstiere nach 20 h. Histopathologie der Gehirne zeigte massive Schädi- gung des zerebralen Cortexes sowie des Hippocampus. Im Gegensatz dazu sanken die bakteriellen Titer der cholinfreien Bakterien nach den ersten acht Stunden mit der Aktivierung des Immunsystems. Alle Tiere überlebten die Infektion unbeschadet.
Durch die Verwendung der gleichen Bakterien im intraperitonealen Maus-Sepsis Mo- dell gelang es mir zu zeigen, dass die cholinassozierte Virulenz unabhänging von Toll-like Rezeptor 2 ist. Obwohl avirulent, waren die cholinfreien Bakterien dennoch in der Lage dendritische Zellen der Milz zu aktivieren und die Produktion von proinflammatorischen Zytokinen und kapselspezifischen Antikörpern zu induzieren.
Diese vorrübergehende Aktivierung des Immunsystems war ausreichend um die cho- linfreien Bakterien aus dem Blutkreislauf zu entfernen sowie eine protektive Immuni- tät auszulösen, welche die Versuchstiere vor folgenden Belastungsinfektionen sero- typspezifisch schützte. Im Gegensatz dazu wuchsen cholinhaltige Bakterien stetig, induzierten eine permante Expression proinflammatorischer Zytokine und führten zum Tod der Versuchstiere.
Von den Experimenten in den Tiermodellen konnte ich darauf schliessen, dass Cho- lin in der Zellwand für das bakterielle Wachstum in Gegenwart einer Immunantwort notwendig ist. Im letzten Teil meiner Doktorarbeit konnte ich zeigen, dass Cholinreste dem Bakterium die Fähigkeit verleihen in komplement-freiem Mausserum ex vivo, oder in Gegenwart des antimikrobiellen Peptids Nisin in vitro zu wachsen. Cholin- freie Bakterien sind dazu nicht in der Lage und werden von einem antimikrobiellen Bestandteil des Serums oder von Nisin beseitigt. Blockiert man die Cholinreste mit cholinspezifischen IgA TEPC-15 Antikörpern oder C-reaktivem Protein (CRP) kann diese Resistenz der cholinhaltigen Bakterien aufgehoben werden. Passive Applikati- on dieser beiden Moleküle kann sogar Versuchstiere vor Besiedlung des Nasopha- rynx sowie vor Sepsis schützen.
1 INTRODUCTION ... 1
1.1 STREPTOCOCCUS PNEUMONIAE... 1
1.1.1 History and Impact on Human health...1
1.1.2 Antibiotic Therapy and Vaccines ...2
1.2 THE PNEUMOCOCCAL SURFACE... 4
1.2.1 The Cell Wall ...4
1.2.1.1 Peptidoglycan ... 4
1.2.1.2 Capsule and Teichoic Acids ... 6
1.2.2 The Role of Surface-bound Choline in the Pneumococcal Physiology ...7
1.2.2.1 Synthesis of Cholinated Teichoic Acids ... 8
1.2.2.2 Choline-binding Proteins... 10
1.2.2.3 Choline-independent Strains ... 11
1.2.3 Surface Proteins of S. pneumoniae ...14
1.3 STREPTOCOCCUS PNEUMONIAE AND THE HOST IMMUNE SYSTEM... 15
1.3.1 Immune response to S. pneumoniae ...15
1.3.1.1 Innate Immunity ... 15
1.3.1.2 Induction of Inflammation... 19
1.3.1.3 Adaptive Immunity... 20
1.3.2 Virulence Factors in Pneumococcal Disease ...21
1.3.2.1 Mechanism of Colonization... 22
1.3.2.2 Mechanisms of Invasive disease... 24
1.4 AIM OF THE WORK... 25
2 RESULTS... 27
2.1 O-ACETYLATION OF PEPTIDOGLYCAN... 27
2.1.1 Identification of Adr as an O-Acetyltransferase using HPLC ...27
2.1.2 Impact of O-Acetylation on Nasopharyngeal Colonization...30
2.1.2.1 Adherence and Invasion of Pen6/Pen6adr to the Pharyngeal Cell Line Detroit 562 ... 31
2.1.2.2 In vitro Growth of R36ASIII and R36ASIIIadr... 32
2.1.2.3 Nasopharyngeal Colonization of R36ASIII and R36ASIIIadr... 32
2.2 CHOLINE RESIDUES OF TEICHOIC ACIDS... 34
2.2.1 Mechanism of Choline-Independence in R6Cho-...35
2.2.1.1 Inactivation of the Wildtype lic2 operon Genes: Impact on Growth and Phenotype of R6Cho-.. 35
2.2.1.2 Choline Content of the Cell walls in lic2 Mutants of R6Cho-... 36
2.2.1.3 Identification of the Inserted S. oralis DNA in Strain R6Cho-... 36
2.2.2 Essential Role of Choline in Pneumococcal Meningitis...39
2.2.2.1 The Pathology of Meningitis is choline-dependent... 39
2.2.2.1.1 Activity score, Virulence and Bacterial Load in the CSF ... 39
2.2.2.1.2 Expression of Matrix Metalloproteinase-9 (MMP-9) in the CSF ... 41
2.2.2.1.3 Histopathology ... 42
2.2.2.2 Inflammation during Meningitis is Choline-dependent ... 44
2.2.2.2.2 Neutrophil Influx into the CSF induced by live Bacteria ... 45
2.2.2.2.3 Cytokine Production in the CSF induced by cell wall preparations ... 46
2.2.2.2.4 Neutrophil Influx into the CSF induced by cell wall preparations... 46
2.2.3 Essential Role of Choline in Pneumococcal Sepsis ...48
2.2.3.1 The Pathology of Pneumococcal Sepsis is Choline-dependent... 48
2.2.3.1.1 Virulence and Bacterial Load in the Blood ... 48
2.2.3.1.2 Toll-like Receptor 2 (TLR-2) does not contribute to the Pathology ... 49
2.2.3.2 Inflammation during Murine Sepsis is Choline-dependent ... 50
2.2.3.2.1 Cytokine Production in the Serum ... 50
2.2.3.2.2 Maturation of Splenic Dendritic Cells ... 52
2.2.3.2.3 In vitro Maturation of human Monocyte-derived Dendritic Cells ... 52
2.2.4 The Role of the Choline residue in Bacterial Growth within the Host...54
2.2.4.1 In vivo growth of choline-free S. pneumoniae in the Murine Host... 54
2.2.4.2 Surface-bound Choline protects S. pneumoniae against the antimicrobial Activity of Murine Serum 54 2.2.4.3 Blocking of Choline Residues and Impact on Pneumococcal Physiology and Deoxycholate- induced lysis ... 56
2.2.4.4 Impact of 50% Choline-content on the Colonizing Capacity of D39Cho- Mutants ... 56
2.2.4.5 Nisin-resistance of S. pneumoniae is Dependent on Surface-bound Choline... 58
2.2.4.6 Choline-specific Immune Molecules protect Mice against Infection with S. pneumoniae... 59
2.2.5 The Protective Potential of Choline-free Strains ...61
2.2.5.1 Induction of Protective Immunity by avirulent Choline-free Pneumococci... 61
2.2.5.2 Production of capsule specific antibodies... 62
3 DISCUSSION ... 64
3.1 O-ACETYLATION OF PEPTIDOGLYCAN... 65
3.1.1 Adr catalyzes the O-Acetylation of the Peptidoglycan...65
3.1.2 The Impact of O-Acetylation on Nasopharyngeal Colonization ...66
3.2 CHOLINE RESIDUES OF TEICHOIC ACIDS... 68
3.2.1 Mechanism of Choline-independence in Strain R6Cho-...68
3.2.2 The Role of Choline Residues in Meningitis and Sepsis...71
3.2.3 The Role of Choline Residues in Immune Clearance of S. pneumoniae ...76
3.2.4 Alternative Host Clearance Mechanisms and Future Projects...82
3.2.5 Protective Potential of Choline-free Strains...85
3.2.6 Future development of live-attenuated vaccine strains ...86
3.3 CONCLUSION... 87
4 MATERIAL AND METHODS ... 89
4.1 MICROBIOLOGICAL METHODS... 89
4.1.1 Cultivation of S. pneumoniae...89
4.2.3 Preparation and Analysis of Cell wall Muropeptides...91
4.2.4 Mass Spectrometry of Muropeptides ...91
4.2.5 Detection of alkaline-labile Acetate in Pneumococcal Cell wall...91
4.2.6 Choline Content of purified Cell walls ...92
4.3 THE INFANT RAT MODEL OF MENINGITIS... 92
4.3.1 The Animal Model ...92
4.3.2 Inoculation of Cell Walls into the CSF space ...93
4.3.3 Cytokine Expression in the CSF...94
4.3.4 Myeloperoxidase Assay ...94
4.3.5 Matrix Metalloproteinase (MMP) Zymography ...94
4.3.6 Histopathology ...95
4.4 THE MOUSE MODELS OF PNEUMOCOCCAL DISEASE... 96
4.4.1 Model of Nasopharyngeal Colonization ...96
4.4.2 Model of Intraperitoneal Sepsis...96
4.4.3 Cytokine Determination in the Serum ...97
4.4.4 Maturation of Murine Splenic Dendritic Cells ...97
4.4.5 Induction of Protective Immunity ...98
4.5 IN VITRO ASSAYS... 98
4.5.1 Maturation of human Monocyte-derived Dendritic Cells...98
4.5.2 Antimicrobial Activity of Murine Serum against S. pneumoniae ex vivo...99
4.5.3 In vitro Killing of S. pneumoniae by the antimicrobial Peptide Nisin...100
4.5.4 Pneumococcal Adherence to the Pharyngeal Cell line Detroit 562 ...100
4.5.5 Detection of capsule-specific IgM antibodies with ELISA ...101
4.5.6 Work with Nucleic Acids and Polymerase Chain Reaction (PCR)...101
4.6 MATERIALS... 102
4.6.1 Instruments, Chemicals, Bacteria, Animals ...102
4.6.2 Culture media ...105
REFERENCES... 109
ABBREVIATIONS... 127
ACKNOWLEDGEMENTS... 131
PUBLICATIONS AND ORAL PRESENTATIONS... 132
EIDESTATTLICHE ERKLÄRUNG ... 134
1 Introduction
Streptococcus pneumoniae (“the pneumococcus”) is one of the first described human pathogens. Despite being known for more than a century all attempts to eradicate this important pathogen have been unsuccessful so far.
Since pneumococcal infections were commonly treated with antibiotics, studies on the pathology of this remarkable microorganism were neglected during the antibiotic era. Only the emergence of antibiotic resistant strains lead to a recurring interest in developing new vaccines that aim to eventually eliminate this bacterium.
To achieve this long term goal it is essential to foster knowledge of the molecular in- teractions between the human host and the pneumococcus. Since the majority of at- tacks launched by the immune system are targeted towards the bacterial surface, a grounding understanding of the bacterial cell envelope and human innate and adap- tive immunity is crucial. Based on this information novel surface-associated virulence factors can be identified and used as potential vaccine candidates to dismantle the bacteria.
1.1 Streptococcus pneumoniae
1.1.1 History and Impact on Human health
S. pneumoniae is a gram-positive bacterium. It was originally isolated and recovered from rabbits infected with human saliva and described simultaneously by George Miller Sternberg and Louis Pasteur in 1881. Initially named Diplococcus pneumoniae the bacterium obtained its present-day’s name Streptococcus pneumoniae in 1974 referring to its phenotypical growth in either diplococcal shape or short filamentous chains.
S. pneumoniae played a pivotal role not only in the discovery of bacterial transforma- tion but also in the designation of DNA being the carrier of the heritable information [1]. After showing that co-injection of an avirulent, but living rough strain with a heat- inactivated, non-viable extract of formerly virulent, encapsulated bacteria into mice resulted in the rapid killing of the animals Griffith concluded in 1928 that these rough
Introduction
minant derived from the heat-inactivated cell extract [1]. Based on these findings Avery further fractionated the crude cell extracts and demonstrated that “nucleic acid of the desoxyribose type is the fundamental unit of the transforming principle”[1]. It was also in the naturally competent S. pneumoniae that “quorum sensing” between bacteria was described for the first time. The synchronized and transient induction of competence in these bacteria is dependent on the concentration of a secreted pneu- mococcal peptide in the growth medium [2].
Normally, S. pneumoniae is a commensal of the human nasopharynx. However, it can become an opportunistic pathogen and therefore is a major cause of mortality worldwide: a WHO report in 2007 estimated that over 1.6 million deaths per year are due to pneumococcal disease. Out of these worldwide fatalities, 0.7-1 million deaths can be attributed to the age group of children below 5 years old [3]. Pneumococcal infection is also one of the main contributors to the severity of viral respiratory dis- ease. A retrospective study indicates that over 50% of the mortality of the 1918 influ- enza epidemic was caused by superinfections with S. pneumoniae [4]. Similarly, the fatality rate of Acquired immunodeficiency syndrome (AIDS) in sub-Saharan Africa is strongly linked to secondary pneumococcal infections [5].
Therefore S. pneumoniae impacts health systems in industrialized and developing countries as well, rendering this bacterium into one of the major global health con- cerns of the past and future.
1.1.2 Antibiotic Therapy and Vaccines
With community-acquired pneumonia being an important human disease the devel- opment of potent pneumococcal vaccines was always of highest priority. After initial vaccination trials using inactivated whole-cell preparations amongst miners in South Africa [6] it was the discovery that purified polysaccharides of the pneumococcal cap- sule are immunogenic and protective themselves [7], which lead to the development of polysaccharide vaccines in the 1940s, covering the capsular serotypes known at that time.
provement of existing vaccines was neglected. The first emergence of clinical multi- ple antibiotic-resistant S. pneumoniae in Johannesburg [8] diverted interest back to the development of advanced vaccines. Although vaccines against 23 of the major capsular types of pneumococci have been introduced in the 1980s [9], they are not effective in children, the primary targets of the pathogen.
For this reason, a heptavalent conjugate vaccine in which 7 important capsular poly- saccharides are chemically linked to an immunogenic protein carrier has been devel- oped and released just recently [6]. Extending the coverage presents increasing technical problems and the cost of the currently available conjugate vaccines is al- ready prohibitive in the very countries where mortality of pneumococcal disease is the highest. In addition, pneumococcal serotypes responsible for most of pneumo- coccal diseases are known to be different between North versus South America, Asia, Africa and Europe.
However, all existing polysaccharide vaccines only comprise a few of the invasive serotypes, are not applicable for children or lead to the selection and emergence of non-vaccine serotypes [10]. Therefore infection with S. pneumoniae still ranks high- est among all vaccine-preventable deaths and the development of alternative pneu- mococcal vaccines is essential. The design of such vaccines implies an elementary knowledge of the molecular processes involved in the pathology of this bacterial dis- ease.
Introduction
1.2 The Pneumococcal Surface
The bacterial surface plays a major role during pneumococcal disease. It represents the interface on which the prokaryotic cell interacts with antimicrobials and the host immune system. Therefore a profound understanding of its architecture and organi- zation is critical to investigate the mechanisms underlying an infection with S. pneu- moniae.
1.2.1 The Cell Wall
A common characteristic of gram-positive bacteria is the presence of a thick cell wall structure, which serves as the outer layer of these bacteria. This sturdy cell wall is mainly built of a gigantic macromolecule called peptidoglycan. Being the physical sur- face of the bacteria the cell wall serves as the anchor for capsular polysaccharides as well as for a variety of surface proteins. It maintains the integrity of the cell against the osmotic pressure and cellular turgor. Despite being the major determinant for cell shape the cell wall still has to retain a certain flexibility since it is involved in key physiological processes such as bacterial growth, cell division, autolysis and the traf- ficking of nutrients.
1.2.1.1 Peptidoglycan
In case of S. pneumoniae the basic structural components of the cell wall are long glycanstrains that consist of the alternating carbohydrates N-acetyl glucosamine (GlcNAc) and N-acetyl muramic acid (MurNAc), the latter of which also serves as the attachment site for the so-called stempeptide, a pentapeptide consisting of L-alanine, D-isoglutamine, L-Lysine and two D-alanine residues (-L-Ala – D-iGln – L-Lys – D-Ala – D-Ala). After minor enzymatic modifications, donor and acceptor stempeptides of adjacent glycan strands can be crosslinked directly or via short, dipeptide bridges containing L-alanine or L-serin (L-Ala – L-Ala or L-Ala – L-Ser), thus forming the basic macromolecular scaffold of the cell wall, called “peptidoglycan” (see Figure 1).
phosphate (UDP)-linked, membrane bound GlcNAc – MurNAc – Pentapeptide pre- cursor, b.) the transfer of Lipid II from the cytoplasm across the membrane onto the bacterial surface, c.) the extracellular extension of the nascent glycan strands with the sugar moieties of Lipid II by a transglycosylation reaction, followed by d.) a transpeptidation reaction that catalyzes the covalent linkage of stempeptides of close proximity between neighboring glycan strands (“crosslinking”). The transglycosylation and transpeptidation reactions are catalyzed by enzymes called penicillin-binding- proteins (PBPs) that are also the major targets of β-lactam antibiotics. Expression of low affinity derivatives of these PBPs confers penicillin resistance to the bacterium and can result in alternative cell wall structures.
Figure 1: The pneumococcal cell wall.
Peptidoglycan strands consist of alternating N-acetyl-glucosamine (GlcNAc) and N-acetyl-muramic acid (MurNAc) molecules. The peptidoglycan strands are linked to each other through stempeptides and short Di-peptide bridges. Wall teichoic acids (WTA) are attached to the MurNAc residues by an unknown linkage. Choline-binding proteins (CBP) bind to the choline residues of the WTA.
Generally, the composition and degree of crosslinking as well as other secondary modifications of the peptidoglycan backbone seem to be highly specific for different S. pneumoniae strains and their features. For instance, the penicillin-resistance of strain Pen6 was highly correlated to the preferential incorporation of the uncommon
Introduction
ucts of alternative alleles in the murMN operon [11]. Also, the deacetylation of GlcNAc components in strain R36A by a peptidoglycan-N-Acetyl-glucosamine- deacetylase (PgdA) enhanced resistance to lysozyme [12].
1.2.1.2 Capsule and Teichoic Acids
The peptidoglycan also functions as the anchor for two important families of carbohy- drates of S. pneumoniae: the capsular polysaccharides and teichoic acids.
S. pneumoniae strains have a vast genetic repertoire for the production of the poly- saccharide capsule which covers the outside surface of these bacteria with one or the other of 91 chemically different capsular polymers [13,14].These highly diverse, chemically distinct polysaccharides are the determinants of pneumococcal serotypes and are the foundation of the nomenclature of S. pneumoniae. Except for the sero- type 3 capsule, all polysaccharides are covalently bound to the cell wall, although the exact site and nature of the bond remains elusive.
In spite of the great variety of capsular polysaccharides the genetic organization en- coding for their synthesis shows a conserved and clustered pattern. Bordered by the flanking genes dexB and aliA the cps locus (capsular polysaccharide synthesis locus) of different pneumococcal serotypes contains genes common to all polysaccharides as well as unique and serotype specific gene sequences [15]. Genes encoding the biosynthesis of chemical components that can be shared with other metabolic proc- esses of the cell (e.g. UDP-GlcNAc) and that can be recruited from common cellular pools are located elsewhere on the bacterial chromosome [15]. Due to the conserved nature of the cps locus, capsular replacement by homologous recombination be- tween two strains can occur and such capsular switching events have been fre- quently observed in vivo [16].
The second important group of surface polysaccharides, present in all pneumococcal serotypes, comprises the structurally related cell wall bound teichoic acid (WTA, for- merly known as “C-polysaccharide”) and the membrane-bound lipoteichoic acid (LTA, formerly known as “Forssmann- / F-antigen”). Due to their characteristic attachment
residues of the peptidoglycan via an unknown linker unit (see Figures 1,2), the incor- poration of LTA into the bacterial membrane is achieved through a terminal glyco- lipid-anchor Monoglucosyldiacylglycerol (Glc-acyl2Gro) (see Figure 2).
Lipoteichoic acid (LTA)
Wall Teichoic acid (W TA)
n n
ChoP ChoP ChoP ChoP
H – Glc – AATGal – GalNAc – GalNAc – Rit – P–
H – Glc – AATGal – GalNAc – GalNAc – Rit – P–
Glc – AATGal – Glc – acyl2Gro (Glycolipidanchor )
? – P – MurNAc (Peptidoglycan)
Figure 2: Structure of Teichoic Acids of S. pneumoniae.
The teichoic acid backbone common to both WTAs and LTAs consists of a basic, repeating subunit, composed of ribitol-5-phosphate (Rit-P), two D-N-acetyl- galactosamines (GalNAc), D-2-acetamido-4-amino-2,4,6-trideoxygalactose (AATGal) and D-glucose (Glc). It was also shown that teichoic acids can be subject to D- alanylation, most likely of the ribitol molecule [17]. A unique and characteristic feature of S. pneumoniae is the presence of the unusual amino-alcohol choline in its cell wall [18]. Depending on the strain, up to two phosphorylcholine molecules can be ester- linked to the two GalNAc residues of the teichoic acid backbone of WTA or LTA, re- spectively [19] (see Figure 2).
1.2.2 The Role of Surface-bound Choline in the Pneumococcal Physiology S. pneumoniae is auxotroph for choline [20], making it an essential nutrient of the growth medium and in vivo environment. The bacterium incorporates choline into its cell wall [18] and attaches it to both the wall teichoic as well as the lipoteichoic acids, thus decorating its surface with this aminoalcohol. Choline serves as the anchor for a
Introduction
family of choline-binding proteins (CBPs) on the pneumococcal surface, which play crucial roles in the bacterial physiology.
1.2.2.1 Synthesis of Cholinated Teichoic Acids
The biosynthesis of cholinated teichoic acids is a co-operative interplay of several parallel enzymatic reactions. While the teichoic acid precursor backbone is intracellu- larly assembled, choline has to be taken up from the extracellular environment, proc- essed within the cell and bound to the teichoic acid precursor backbone, before the cholinated teichoic acid can be flipped across the membrane and connected to the peptidoglycan scaffold of the cell wall. In contrast to the fully understood choline me- tabolism the production of the teichoic acid backbone is only partially known. So far eight genes have been described to participate in the synthesis reactions. These genes are clustered in two genetic loci, designated lic1 and lic2 operon (see Figure 3).
Figure 3: Genetic organization of the lic1 and lic2 operons.
Both operons are involved in the utilization of choline (genes licA, licB, licC, licD1/2) and the synthesis of teichoic acids (genes tacF, tarI, tarJ).
The only identified genes contributing to the production of the intracellular teichoic acid precursors (in particular the ribitol phosphate subunit) are located in the lic1 op- eron [21]. TarJ, a NADPH-dependent alcohol dehydrogenase catalyzes the synthesis of ribitol 5-phosphate from ribulose 5-phosphate. Subsequently, the activation of ribi- tol 5-phosphate to cytidine 5’-diphosphate (CDP)-ribitol is achieved by the cytidylyl- transferase TarI [21]. However, it is not known how CDP-ribitol is incorporated into the teichoic acid precursor molecule nor how the remaining TA backbone is synthe-
TacF
P-Cho CDP-Cho Cho
ChoP ChoP
CDP
PCho PCho
LicA
LicC
Peptidoglycan
Membrane Cytosol
LicD1 LicD2
TA Precursor
TarJ/I Encoded in lic1 operon
LicB Cho
Cho Cho
Cho Cho
Encoded in lic2 operon
Figure 4: Synthesis of Cholinated Teichoic Acids (TA).
Extracellular choline is taken up by a choline transporter LicB, phosphorylated by an ATP-dependent choline kinase LicA, and activated by a phosphorylcholine cytidylyl transferase LicC. The phosphoryltransferases LicD1/D2 catalyze the attachment of choline to the TA acid precursor. Upon full cholination, the TA is transferred across the membrane by flippase TacF. The mechanism of attachment to the peptidoglycan scaffold is not known.
Proteins TarJ/I are involved in the synthesis of the TA precursor.
In parallel, the acquisition of extracellular choline from the growth medium and its cy- toplasmic processing has to take place. This is catalyzed by genes encoded in both the lic1 and lic2 operon. After choline uptake by a transporter LicB , and intracellular phosphorylation of the aminoalcohol by an ATP-dependent choline kinase LicA [22,23], the resulting phosphorylcholine is further activated to CDP-choline by the phosphorylcholine cytidylyl transferase LicC [22,24,25], in a CTP-dependent manner.
The first three enzymes LicA/B/C are encoded in the lic1 operon. Following these enzymatic reactions, gene products of the lic2 operon catalyze the loading of teichoic acid chains with choline residues: each of the phosphorylcholine transferases LicD1 and LicD2 [26,27] attaches one CDP-choline derived phosphorylcholine residue to the two GalNAc residues of the TA precursors.
Once the cholination of the TA precursors is accomplished flippase TacF transports the choline-containing TAs onto the bacterial surface [28], where they are linked to the bacterial cell wall by a presently unknown mechanism (see Figure 4).
Introduction
Teichoic acids were also described to be further modified: for instance the transfer of D-alanyl residues to the backbone is dependent on the presence of a functional dlt operon in the bacterial genome [17]. Interestingly, an extracellular phosphoryl-choline esterase Pce can secondarily remove choline-residues from TAs [29].
1.2.2.2 Choline-binding Proteins
Choline in the cell wall can dramatically influence physiological properties of S.
pneumoniae. While the unique auxotrophic requirement for choline can be fulfilled by other structurally different amino-alcohols [30], the normal physiological properties of the bacterium require the trimethylamino group of choline. S. pneumoniae growing in media in which choline was replaced by ethanolamine show numerous abnormalities:
they form long chains, do not autolyse and cannot undergo genetic transformation [31]. These fundamental phenotypical changes of the bacterium can mainly be attrib- uted to the role of choline residue as an anchor for a versatile family of non- covalently bound choline-binding proteins (CBPs) (see Figure 1).
Besides their role in virulence (as discussed below) some CBPs contribute to the cell biology of the pneumococcus, especially to the physiology of the cell envelope. A common structural feature of CBPs is the existence of a C-terminal choline-binding domain [32]. Screening of available S. pneumoniae genomes for potential choline- binding domain sequences suggests the presence of 10 (in strains R6, D39) to 15 (in strain TIGR4) CBPs on the bacterial surface [33].
The first characterized and most intensively studied CBP is LytA, a N-acetyl- muramoyl-L-alanine-amidase, which is responsible for the phenomenon of autolysis.
After reaching the stationary phase cultures of S. pneumoniae undergo self-induced lysis in a choline-dependent manner [31].
The physiological “purpose” of this process is not understood yet. The presence of LytA is also required for penicillin- and deoxycholate-induced lysis of the bacteria.
Another CBP LytB, a β-N-acetylglucosamidase, cleaves the glycan strands and is responsible for daughter cell separation [34]. LytB mutants or bacteria grown in the
autolytic activity of which is regulated by the amino-terminal domain of another re- cently crystallized protein CpbF [36]. CBPs can also have regulatory effects on the transformation process of competent pneumococci as shown for CbpD, a putative murein hydrolase, which is supposed to be responsible for competence-induced cell lysis and DNA release [37].
Attachment of CBPs to the bacterial surface is dependent on the availability of vacant choline residues on the surface. A conceivable superior control mechanism on CBP attachment could therefore be the regulation of vacant choline-residues by another choline-binding protein Pce. Being a phosphoryl-choline esterase, Pce can remove phosphorylcholine residues from TAs [29].
1.2.2.3 Choline-independent Strains
Interestingly, the aforementioned ethanolamine induced phenotypical changes (e.g.
chain growth, autolysis-deficiency) are exactly the same abnormalities shown by all known isolated choline-independent strains of S. pneumoniae when cultured in cho- line-free media [28,38,39] (see Figure 5).
Most of the studies on choline were performed on two choline-independent strains (Cho-, JY2190), which were the only available and existing mutants for years. Just recently, several additional new mutants were obtained [28,40]. Most of these cho- line-independent strains were laboratory mutants isolated by an enrichment proce- dure in which the parental strain was serially passaged in a culture medium, the cho- line component of which was replaced by gradually decreasing concentrations of ethanolamine.
Introduction
Figure 5: Morphology of R6Cho- and R6Cho-licB in choline-free and choline-containing media [41].
Both strains are able to grow in choline-free medium in long, autolysis-resistant chains. After supplying choline to the growth medium R6Cho- still has the ability to utilize choline and revert to the diplococcal wildtype phenotype.
In contrast, mutants deficient in any of the choline utilization genes of the lic1 operon (e.g licB) will maintain the choline-free phenotype even in the presence of exogenous choline.
This procedure eventually yielded mutants R6Chi [28], JY2190 [39] and a whole fam- ily of strains ranging from P023-P600 [40] which could grow in media completely lack- ing the aminoalcohol component. The first genetic and biochemical studies were per- formed on R6Chi and identified a single GT point mutation in one of the genes of the lic2 operon as the molecular basis of choline independence in this mutant [28].
During these studies, the gene tacF was discovered and proposed to encode for a polysaccharide transmembrane transferase (“flippase”) that catalyzes transport of teichoic acid chains to the outer surface of the pneumococcal plasma membrane [28]
(see Figure 6).
In contrast to the wildtype TacF that only transfers TA precursors with aminoalcohol moieties, it appears that the mutated form allows the transfer of unsubstituted precur- sors, thus enabling the survival of the bacteria (see Figure 6).
Analyzing strains JY2190 and P023-P600 revealed that differently located mutations in tacF can also confer choline-independence to the bacterium. So far neither the ex- act nature of the needed aminoacid alteration nor the required changes in structure
Cytosol
R6 R6Chi
WT TacF mutant TacF
TA precursor Peptidoglycan
ChoP ChoP
PCho PCho
Membrane
Figure 6: Mechanism of choline-independence in the Chi background.
In wildtype R6 the flippase TacF (left side of the panel) is restricted to transferring fully cholinated teichoic acid (TA) precursors to the surface, where they are attached to the cell wall. In contrast, flippase TacF of strain R6Chi (right side of the panel) possesses a mutation that also allows the transport of choline-free TA precursors across the membrane.
However, it seems that a modified TacF is the crucial element in delivering choline- independence to these strains.
The mechanism of choline-independence in the other extensively studied S. pneu- moniae strain R6Cho- [38] was not known at the beginning of my thesis. R6Cho- shared many of the physiological properties of R6Chi / JY2190 / P023-P600 but had a more complex origin: it was isolated as the product of a heterologous genetic cross in which the recipient S. pneumoniae strain R6 was transformed with donor DNA from S. oralis, a streptococcal species that contains choline in its teichoic acid but has no auxotrophic requirement for it [38]. The resulting transformant R6Cho- lost its auxotrophic need for choline, and was the bacterial strain predominantly studied in this work.
All choline-independent strains will maintain the wildtype phenotype whenever cho- line is available. Choline utilization can be avoided by deletion of any of the genes in the lic1 operon (e.g. licA, licB, licC) (see Figure 5).
Introduction
1.2.3 Surface Proteins of S. pneumoniae
The pneumococcus has numerous surface proteins that can be divided into four ma- jor groups, depending on their secretion pathways and the nature of the chemical bond linking them to the surface.
The first group comprises proteins that are covalently attached to the peptidoglycan.
These proteins contain a common c-terminal LPxTG-motif, which is recognized by Sortase A (SrtA), a transpeptidase that catalyzes the covalent linkage of these pro- teins to the peptidoglycan [42]. It was shown that the sortase-dependent proteins neuraminidase A (NanA) and β-galactosidase (β-Gal) were released into the growth medium by srtA mutants of the R6/D39 background. Additional sortase genes (srtB, srtC, srtD) were found in strain TIGR4 associated with the strain’s virulence [42].
A second group of surface proteins - only found in pneumococci in such abundance – are the aforementioned choline-binding proteins (CBPs), named after their non- covalent binding to the unusual choline-residue of the cell wall (see above).
Lipoproteins of S. pneumoniae include proteins that function in mechanisms as di- verse a manganese-transport and adhesion (PsaA, pneumococcal surface adhesion A), iron-uptake (PiaA and PiuA) or as chaperons belonging to the peptidyl-prolyl isomerase family (PPIases) [43].
Additional proteins that cannot be assigned to any of the three above mentioned, classical groups are summarized as “moonlighting” proteins [43], but will not be dis- cussed in further detail here.
1.3 Streptococcus pneumoniae and the Host Immune System In the majority of cases S. pneumoniae primarily colonizes the mucosal surface of the nasopharynx in a transient and asymptomatic manner. However, for unknown rea- sons, bacteria can occasionally descend to the lungs and induce potentially lethal pneumonia. A dysfunctional Eustachian tube can also grant access to the middle ear, leading to otitis media [44]. Most likely originating from the lungs, the bacteria can disseminate into the bloodstream and lead to systemic bacteraemia or even lethal septicemia. The most severe outcome of pneumococcal disease is the infection of the brain and meningitis.
Mucosa, bloodstream and brain are immunologically very different host compart- ments. In all these in vivo environments S. pneumoniae has to defend itself against versatile immune effectors. To ensure survival, the bacterium is equipped with an arsenal of virulence factors to react and adjust to these host defense strategies.
The following chapter will describe the major modes of infection and focus on se- lected “host-pneumococcus-interactions”, particularly important to understand the data presented in this work.
1.3.1 Immune response to S. pneumoniae
Upon infection of the human host S. pneumoniae is the target of a variety of defense mechanisms. The immediate host measures to pneumococcal infection involve pathogen-unspecific immune effector mechanisms of the innate immunity that seek to eradicate the bacteria.
The simultaneous induction of inflammation helps to amplify this immune response.
Inflammation also jumpstarts the host’s adaptive immunity that will eventually kill the invaders and lead to a (long-term) pathogen-specific immunological memory.
1.3.1.1 Innate Immunity
Pneumococci trying to infect the human host have to successfully overcome innate immunity, which constitutes the first, intrinsic line of defense. The germ-line encoded
Introduction
innate immune effector mechanisms found on the mucosa are different from the one’s present in the bloodstream. Nevertheless, they are always directed against conserved molecular structures that are abundant in a multitude of pathogens.
For example, an effective defense can already be achieved by physical barriers.
Cells of mucosal epithelia are sealed by tight junctions and secrete viscous mucus to avoid colonization and invasion. The mucus covers and agglutinates invading bacte- ria which will subsequently be expelled by the movement of the cilia. The existing commensal microflora of the target tissues competes with intruding pathogens for nutrients and physical space inhibiting their attachment. Another very effective mechanism to avoid bacterial proliferation is to establish a pH barrier or to shift the temperature away from the bacterium’s growth optimum, which occurs during in- flammation or fever [45].
Besides these straightforward but simple defense strategies, innate immunity also possesses manifold sophisticated molecular strategies to tackle bacteria.
For instance, lysozyme can be found in saliva and mucus and is especially effective in the clearance of gram-positive bacteria, as it directly attacks the cell wall and the stability of the whole bacterial cell itself. Being an N-acetylmuramide glycanhydrolase it cleaves the β(14) glycosidic bond between the GlcNAc and MurNAc residues of the peptidoglycan, resulting in the enzymatic degradation of the cell wall and the lysis of bacteria [46].
Another way to fight bacteria is the secretion of cationic antimicrobial peptides (CAMPs). Studies in human patients as well as in animal models demonstrated ele- vated levels of these peptides during bacterial mucosal colonization, septicemia and meningitis [47,48,49,50,51]. It is believed that the peptides’ mode of action is mainly mediated through their cationic charge, which attracts them towards the anionic phospholipids of bacterial membranes. Due to their hydrophobic nature the peptides are able to disrupt the membrane and eventually kill the pathogen [52].
Also found on mucosal surfaces is the secretory, dimeric immunoglobulin A (IgA). IgA is produced in high abundance, and is assumed to play a crucial role in the protection of the host against mucosal pathogens [53]. However, the exact effector functions of
herence of bacteria to mucosal epithelia or neutralizes bacterial toxins [45]. A mono- meric form of IgA is found in the serum. Since IgA is only a weak inducer of comple- ment (see below) it is concluded that serum IgA has to have additional, unknown mi- crobicidal effects [53]. Interestingly, naturally occurring choline-specific serum IgG and IgM antibodies were shown to be present in murine and human serum and can be protective against pneumococcal infection [54].
Once in the tissues or the bloodstream, bacteria have to face additional powerful de- fense mechanisms.
A very well characterized effector is “complement”. The major goals of this compli- cated system are to directly kill pathogens, to coat (“opsonize”) the microbial surface and to induce inflammation at the site of infection (see below). In general, opsoniza- tion with immune molecules (e.g. complement, antibodies…) facilitates the ingestion of pathogens by macrophages and neutrophils. The activation of complement can be initiated through the recognition of bacteria by complement complex C1q in combina- tion with IgG and IgM (“classical pathway”), by Mannose-binding lectin (MBL) that preferably recognizes mannose or GlcNAC (“lectin pathway”) [55,56] or via sponta- neous binding of complement on the pathogen’s surface (“alternative pathway”) [45].
In contrast to pathogens, host cells are protected against the spontaneous deposition of complement proteins induced by the alternative pathway due to presence of cer- tain surface proteins, such as factor H, which displaces erroneously bound comple- ment factors. Despite the differences in activation, all three pathways merge at the formation of a C3 convertase that converts the serum factor C3 into the inflammatory C3a molecule and C3b, the latter of which eventually opsonizes the whole pathogen.
Serum factors C5b-C9 form the so-called membrane-attack complex (MAC) and lead to membrane perforation and subsequent killing of bacteria. However, due to its thick peptidoglycan layer S. pneumoniae is protected against the microbicidal effects of the MAC [57].
Another immune molecule contributing to the clearance of S. pneumoniae is the C- reactive protein (CRP). It is a liver-synthesized acute phase serum protein, rapidly secreted during inflammation. Although it was originally discovered in the serum of patients suffering from pneumococcal pneumonia [58], later studies also demon- strated its presence in the upper respiratory tract and in nasal secretions [59]. C-
Introduction
reactive protein was named after its affinity for WTA (formerly C-polysaccharide). Fur- ther investigations revealed that CRP has a high specificity towards the phosphoryl- choline residue of either WTA or LTA of S. pneumoniae [60]. Just recently another choline-specific pentraxin Serum amyloid P (SAP) was described [61]. Using direct competition assays it was determined that the overall avidity of SAP towards the cho- line residue was lower than the one of CRP [61]. Both, CRP and SAP can opsonize and activate complement via the classical pathway as well as induce phagocytosis by macrophages and neutrophils [62,63,64]. It was shown that the presence of SAP or heterologous human CRP is important to control pneumonia and bloodstream infec- tions of S. pneumoniae in mice [62,65].
Innate immunity also possesses several cellular defense mechanisms. After crossing the epithelial barrier and infecting submucosal tissues, S. pneumoniae will encounter resident phagocytic host cells, called macrophages. In the course of an infection more phagocytic white blood cells, the neutrophils, will evade from the bloodstream and influx into the site of inflammation. These phagocytes express numerous surface receptors which recognize bacterial components or previously opsonized pathogens and trigger the engulfment and ingestion of the microorganisms. The macrophage mannose receptor was shown to bind to pneumococcal capsular polysaccharides, although the exact epitope on the bacterial polymers could not be determined yet [66]. Scavenger receptors, which represent another receptor family, target negatively charged LTA or peptidoglycan components [67]. In particular the macrophage recep- tor with collagenous structure (MARCO) was shown to play a crucial role in the clear- ance of S. pneumoniae by alveolar macrophages during pneumonia [68]. Similarly, the removal of pneumococci by macrophages of the marginal zone of the spleen de- pends on the presence of a C-type lectin receptor named SIGN-R1 [69]. Complement and Fc receptors facilitate the uptake of opsonized microorganisms through the rec- ognition of bound complement factors or antibodies, respectively.
Upon internalization bacteria are contained in a vesicular phagosome, which - upon fusion with lysosomes - becomes a phagolysosome. Within this compartment bacte- ria are exposed to a whole array of microbicidal mechanisms, ranging from acidifica- tion, enzymatic degradation or nutrient deprivation to cationic antimicrobial peptides.
A recently discovered altruistic defense mechanism of neutrophils is the expulsion of neutrophil extracellular traps (NETs). Hereby, neutrophils eject DNA which forms a web of chromatin that traps and kills bacteria together with other released bactericidal agents [70,71].
1.3.1.2 Induction of Inflammation
In parallel to the innate immune system’s effort to eradicate attacking pathogens im- mediately, the host initiates an inflammation which will orchestrate a complex inter- play of several immune mechanisms in order to amplify the overall immune response.
Dendritic cells (DCs) and macrophages are the major modulators of inflammation.
Expressing pattern recognition receptors (PRRs) they permanently sense their sur- rounding environment and the phagocytosed materials for the presence of pathogens by scanning for highly conserved pathogen-associated molecular patterns (PAMPS).
For instance, the family of Toll-like receptors (TLRs) comprises ten members, out of which TLR-2 binds to pneumococcal LTA, TLR-4 to the major pneumococcal toxin pneumolysin (see below) [72] and TLR-9 to unmethylated CpG motifs specific to bac- terial DNA [73]. The intracellular protein Nucleotide-oligomerization domain 2 (Nod 2) receptor detects cytosolic muramyl-dipeptides of S. pneumoniae, a cell wall compo- nent consisting of muramic acid and L-Ala and D-iGln residues [74]. Two members of a recently discovered family of Peptidoglycan-recognition-proteins (PGRPs), the ami- dase PRGP-L [75] and PGRP-S, were shown to contribute to S. pneumonia recogni- tion and clearance although the exact mechanism could not be clarified yet [76].
Upon detection of intruding pathogens, DCs are activated and start to secrete proin- flammatory cytokines, called Interleukin 1β (IL-1β), IL-6, Tumor Necrosis Factor alpha (TNF-α), CXCL8 (formerly IL-8) and IL-12. Especially IL-1β, IL-6 and TNF-α mediate local inflammation characterized by the constriction of adjacent blood vessels, the activation/permeation of the vessel’s endothelia and the detour of the blood/lymph flow to the infection site. A similar local inflammatory response can also be triggered by small molecules of the activated complement cascade such as C3a, C4a and C5a.
This inflammation process alleviates the influx of innate effector molecules (comple- ment, antibodies) or macrophages and neutrophils into the site of inflammation.
Introduction
The consecutive interaction of adherence molecules on both, the activated vascular endothelium (e.g. Selectins, ICAM-1) and circulating, activated neutrophils (e.g. In- tegrins), leads to a first reversible “rolling” of neutrophils on the endothelium followed by the tight binding to each other. Attached neutrophils cross the basement mem- brane and migrate within the tissue along the chemical gradient of chemokine CXCL8 towards the infectious focus, where they can execute their bactericidal effects.
To assure the constant supply of reinforcing immune cells IL-1β, IL-6 and TNF-α also have systemic effects and stimulate the production of neutrophils in the bone marrow.
These cytokines also regulate the body temperature and lead to fever, which aims to inhibit the pathogens’ growth. They also induce the acute phase response with the production of CRP, SAP [62] and MBL in the liver. These molecules can trigger com- plement and phagocytosis by neutrophils. CRP was also shown to elicit IL-1β and TNF-α secretion in vitro [58].
Another effect of TNF-α is to stimulate the drainage of dendritic cells within the lymph fluid flow towards the lymph nodes where DCs mature and initiate an adaptive im- mune response [45].
1.3.1.3 Adaptive Immunity
Immature dendritic cells aim to detect pathogens with PRRs. Then they enzymatically disassemble proteins of the intruder and present the digested pathogen-derived pep- tide components on surface molecules named major histocompatibility complex I or II (MHC-I, MHC-II). Simultaneously the DC secretes cytokines (see above), stops phagocytosis and MHC turnover. Thus it “freezes” its surface and presents these short pathogen-derived peptide fragments. Activated DCs also upregulate costimula- tory molecules CD80 and CD86 and start migrating to the lymph nodes [45].
In the lymph node they get in close proximity to naïve T-cells. The T-cell receptor of each T-cell clone has a certain specificity and can bind to surface MHC molecules on DCs. Once a T-cell specifically recognizes a pathogen-peptide-MHC complex, and additionally receives a second signal from the costimulatory molecules CD80 or
In the case of a protein-specific adaptive immune response against the extracellular pathogen S. pneumoniae the differentiated effector T-cells activate antigen-specific B-cells. These B-cells become antibody-secreting plasma-cells that can potentially establish a long-term immunological memory. The peptide-specific antibodies aim to opsonize bacteria, stimulate complement deposition or interfere with the pathogen’s attachment to target tissues. Antibodies to pneumococcal proteins PsaA, CbpA and PspA were shown to be protective against nasopharyngeal colonization [77,78,79].
On the other hand, production of naturally occurring antibodies specific to the capsu- lar polysaccharide and phopshorylcholine components of S. pneumoniae appears to be T-cell-independent (TI). A certain subpopulation, called B-1 cells, can directly be activated upon contact with these pneumococcal TI antigens and contributes to the rapid secretion of circulating antibodies of the IgM or IgG isotypes. However, a T-cell independent immune response does not lead to a long term immunological memory [80].
These two described immune mechanisms are also the crux of the efficacy of pneu- mococcal vaccines. A protein specific vaccine would be able to induce long-term im- munity in the host. Yet, no such immunogenic pneumococcal protein compound common to all strains and serotypes has been identified that satisfactorily induces immunity. In contrast the known polysaccharide vaccines consist of TI antigens and do not establish strong and powerful long-term protection. A recently released poly- saccharide vaccine Prevnar tries to overcome these problems by chemically coupling the polysaccharide to the cholera toxin protein thus triggering an immunological memory. Unfortunately, the serotype coverage of this vaccine is limited to the re- stricted amount of polysaccharide molecules that can be attached to the protein car- rier molecule.
1.3.2 Virulence Factors in Pneumococcal Disease
S. pneumoniae is very flexible and infects various host tissues and organs. To adapt to these various growth environments, pneumococci can regulate the expression of their capsule, a process named “phase variation” [81] (see Figure 7). Similarly, the bacteria switch between two growth modalities, planktonic life in the bloodstream and
Introduction
biofilm growth on lung tissues, in the middle ear (otitis media) and during meningitis [82]. Both - the two phases as well as the two growth modalities - are characterized by specific gene expression profiles including multiple virulence genes.
1.3.2.1 Mechanism of Colonization
Pneumococcal disease starts with the colonization of the host. This process can be divided into two sequential components: the bacterium has to evade innate immunity of the mucosa first, before multiple receptor-ligand interactions promote adherence to nasopharyngeal and lung epithelial tissues.
A major virulence factor of S. pneumoniae is the pneumococcal capsule, which fulfills specific tasks in respect to the site of infection. Its role in the mucosal compartment is to reduce the agglutination of pneumococci by mucus and to limit the bacterium’s ejection from the host system [83]. To inhibit the hydrolytic and bactericidal effects of the abundantly present lysozyme the bacterium expresses PgdA, a protein that en- zymatically N-deacetylates the GlcNAc units of the peptidoglycan strands and con- fers resistance [12]. S. pneumoniae has also to compete with the residential, com- mensal microflora of these organs to create its own growth niche. Interestingly, it was demonstrated that the production of hydrogen peroxide by S. pneumoniae efficiently displaces Staphylococcus aureus from the nasopharynx [84] allowing S. pneumoniae to settle. To repel secreted cationic antimicrobial peptides from its membrane the pneumococcus attaches positively charged D-alanyl residues to its teichoic acids, thus lowering not only the net negative charge of the surface but also its susceptibility towards CAMPs [17].
After bypassing the immune system S. pneumoniae possesses several molecular adhesion strategies, with which the bacterium is able to colonize host epithelia.
Pneumococcal phosphorylcholine residues show structural homology to the host molecule Platelet-activating factor (Paf), thus promoting binding to its respective re- ceptor (rPaf) on the epithelial cells of the mucosa [85]. The pneumococcus even uses the host cell’s recycling pathway of the rPaf to cross the epithelial barrier and enter
tory system [32]. Choline-binding protein A (CbpA) also functions as an adhesion molecule of pneumococci recognizing polymeric Ig receptors on host cells [43].
Besides choline-associated binding mechanisms S. pneumoniae has the ability to interact with glycoconjugates on host cells. For example, the addition of N- acetylglucosamine–β-1-3-galactose (GlcNAcβ13Galβ) inhibits pneumococcal bind- ing to human pharyngeal cells in vitro [86]. Similarly, other oligosaccharides were de- scribed that promote attachment of pneumococci to epithelial cells [87,88,89].
The recently discovered pneumococcal pilus was also suggested to contribute to host cell adherence and virulence of pneumococci [90]. However it does not seem to be a major adhesin of S. pneumoniae since screening of different pneumococcal sero- types revealed that the pilus genes are present only in the minority of pneumococcal strains [91].
Although it has protective effect against the mucus, the polysaccharide capsule of the bacteria represents a major problem for the adherence process, since it masks and shields the underlying cell wall localized adhesion molecules needed for attachment to epithelia. Therefore the surface of the pneumococcus varies between two different growth phases, a “transparent”, colonizing one, in which capsular polysaccharide content is the lowest and expression of cholinated teichoic acids and adhesion mole- cules (e.g. CbpA) is high, and an “opaque”, invasive phenotype, in which the bacte- rium is protected by a thick capsule against defense mechanisms occurring upon in- vasion [81] (see Figure 7). It also appears that the capsule interferes with the formation of biofilms, the preferred growth modality of S. pneumoniae during coloni- zation [82,92]. Actually, the common upregulation of some genes (e.g. nanA) found in expression profiles of transparent, colonizing colonies and biofilm-derived cells might suggest a correlation between these two modalities of growth [82].
